Complete local text-to-waveform speech synthesis under 10M parameters. Fixed-voice English TTS with deterministic seeds, long-text handling, and CPU or CUDA inference.
I built and funded Inflect v2 independently. If this release finds a real audience, I would like to continue the project with a broader v3, which might include things like more langauges, voices, and stability improvements. If the model is useful to you, leaving a like on Hugging Face genuinely helps more people discover it.
Inflect v2 uses one public API across two sizes: Micro prioritizes quality below 10M parameters; Nano prioritizes footprint below 4M.
These are held-out text generations, not reconstructions of training audio. Each transcript is shown exactly as passed to the public frontend.
No single metric captures TTS quality. Inflect v2 reports human preference , predicted naturalness , multi-ASR intelligibility , complete footprint , and runtime separately rather than compressing them into one unverifiable score.
The headline row always refers to Inflect-Micro-v2 . Detailed competitor results and protocol boundaries are kept visible below.
Comparison set. Results include KittenTTS Nano：https://huggingface.co/KittenML/kitten-tts-nano-0.8, Piper Low：https://huggingface.co/rhasspy/piper-voices, and Supertonic 3：https://huggingface.co/Supertone/supertonic-3, established compact or local TTS baselines with larger deployable weight footprints than both Inflect releases. Weight sizes are compared at package level, and no single metric is treated as proof of overall superiority.
Inflect-Micro-v2 recorded a 66.2% preference rate (21 wins · 10 losses · 3 ties) in the final anonymous community study. Systems were hidden, left/right order was randomized, and ties count as half a win. This is descriptive community evidence, not formal MOS.
The UTMOS22 run used 500 identical unseen prompts per voice. KittenTTS and Piper are equal-weight two-voice means; their observed voice ranges appear as whiskers. Supertonic 3-step is reported below the plotted range rather than flattening every other system.
Inflect-Micro-v2: 4.395 UTMOS22 , 95% bootstrap CI 4.381–4.408 . UTMOS22 is a learned predictor, not human MOS.
The headline score is the equal-weight mean of Qwen3-ASR and Nemotron 3.5 corpus WER for every system. Whisper is excluded consistently from the headline because it produced insertion-heavy hallucinations on a subset of otherwise intelligible Supertonic 8-step clips. It is not deleted: the complete three-ASR evidence remains below.
For Inflect-Micro-v2, the individual results are 2.52% Qwen3-ASR , 5.45% Nemotron 3.5 , and 2.73% Whisper large-v3 . The former three-model mean, 3.57% , is retained only as a descriptive audit value and is not used as the headline score.
These views are diagnostics, not additional leaderboards. They show where the recognizers disagree and which prompt categories still produce recoverable transcription errors.
Both Inflect releases synthesize comfortably faster than real time on CPU. The managed reference run used a Hugging Face CPU Upgrade instance (8 vCPU, 32 GB RAM) with four framework threads , end-to-end text-to-waveform timing, and 100 fixed Modern400 prompts. Three complete passes were recorded; the first cache-building pass was excluded and the table pools passes two and three.
These are package-level results from the public PyTorch runtime, not a claim that Inflect is the fastest compact TTS system. Hardware, frontend behavior, framework, compilation, and thread policy all affect small-model measurements.
The same managed CPU and four-thread policy were used for a shorter comparator pass: the identical 50-prompt prefix, repeated twice. KittenTTS and Piper are equal-work pooled across their two tested voices.
Because Inflect uses the larger 100-prompt steady-state run while comparator rows use the shorter 50-prompt confirmation pass, this table is deployment context rather than a perfectly matched speed leaderboard. Several comparators also use optimized ONNX runtimes, while the published Inflect benchmark above uses the canonical PyTorch runtime. The separately released Inflect ONNX path has not been substituted into those benchmark numbers.
Voice variants sharing the same weights are merged. Inflect totals include the integrated waveform decoder.
Inflect-Micro-v2 is the quality-focused member of the family. Both models use the same public API and complete text-to-waveform packaging.
This uses the Hub's version-aware downloader and retrieves the complete repository. A Git clone also works, but hf download is the recommended path for ordinary model installation.
The result is a 24 kHz mono float32 waveform. Long input is split at punctuation-aware boundaries, synthesized chunk by chunk, and joined with controlled pauses.
The official verified FP32 export is published separately as Inflect-Micro-v2-ONNX ：https://huggingface.co/owensong/Inflect-Micro-v2-ONNX. It supports dynamic lengths, CPU/CUDA/DirectML provider selection, deterministic seeds, and the same long-text wrapper without importing PyTorch:
The neural model is split into duration.onnx and decode.onnx ; together they contain the complete learned text-to-waveform path. The English eSpeak-ng frontend remains CPU-side code. See the ONNX repository ：https://huggingface.co/owensong/Inflect-Micro-v2-ONNX for graph contracts, provenance, parity measurements, browser deployment, and re-export instructions.
Inflect v2 is a parameter-efficient VITS-family end-to-end text-to-waveform generator with an English phoneme frontend, monotonic alignment, stochastic latent synthesis, residual coupling flow, and an integrated alias-reduced neural waveform decoder.
The release describes the deployable architecture. Private corpus-construction and optimization details are not part of this open-weight package.
Long passages are punctuation-aware chunks, not one unlimited autoregressive pass. Chunk boundaries receive short pauses and edge fades. See docs/API.md ：https://huggingface.co/owensong/Inflect-Micro-v2/blob/main/docs/API.md for waveform contracts and concurrency notes.
The release contains one fixed synthetic English voice. The package does not redistribute a real-speaker recording corpus, does not claim the voice as the identity of a real person, and requires no reference audio or external model at inference.
This release is inference-first. New-voice and new-language adaptation are not currently validated or supported . A new voice would replace the fixed speaker rather than add a selectable speaker; language adaptation also requires rebuilding normalization, phonemes, symbols, embeddings, and training data. See docs/DATA_AND_VOICE.md ：https://huggingface.co/owensong/Inflect-Micro-v2/blob/main/docs/DATA_AND_VOICE.md and docs/FINETUNING.md ：https://huggingface.co/owensong/Inflect-Micro-v2/blob/main/docs/FINETUNING.md.
Do not use the included voice to impersonate a real person, deceive listeners, or create fraudulent content. Disclose synthetic speech where the context could otherwise mislead. Users are responsible for applicable laws and the Apache-2.0 license.
Original Inflect code and weights are released under Apache-2.0. Bundled third-party components retain their own notices in THIRD_PARTY_NOTICES.md ：https://huggingface.co/owensong/Inflect-Micro-v2/blob/main/THIRD_PARTY_NOTICES.md. release_manifest.json records packaged file sizes and SHA-256 hashes.
Inflect v2 is an open-weight release. Deployable weights, inference code, frontend code, evaluation prompts, and release reports are public. The training corpus-generation pipeline, private filtering infrastructure, and full optimization recipe are not part of the public package.
Owen Song may share additional technical context privately for credible research, collaboration, reproducibility, or deployment inquiries when the request has a clear purpose and does not conflict with licensing or data-provenance constraints.
Designed and developed independently by Owen Song · open weights · Apache-2.0 · complete local text-to-waveform inference
扫码直达文章详情